You need to sign in or sign up before continuing.
Back to History

Fields Changed

Registration

Field Before After
Trial Title How Leaders Emerge: Evidence from Gender Composition, Leadership Selection, and Team Performance in India How Leaders Emerge: Evidence from Leadership Selection and Team Performance in India
Abstract Jobs that involve non-routine analytical team tasks have grown rapidly over the past two decades, increasing demand for soft skills essential to these roles. My study examines the interplay of gender composition, teamwork, and leadership in the context of higher education and STEM in India. Specifically, I propose an experiment with 610 students at a large engineering college, where students are randomly assigned to project teams that vary along two dimensions: team gender composition (female-majority vs. female-minority teams) and leadership selection method (peer-nominated vs. researcher-assigned leaders based on social skills). Using an incentivized app development competition addressing challenges in rural India, the study analyzes how gender composition affects leadership emergence, examines differences in leader characteristics across selection methods, and measures how these factors individually and jointly influence team dynamics and performance. The study allows me to investigate both the role of gender composition in team leadership and performance and the effectiveness of different leadership selection mechanisms in a developing country context. Jobs that involve non-routine analytical team tasks have grown rapidly over the past two decades, increasing demand for soft skills essential to these roles. My study examines the interplay teamwork and leadership in the context of higher education and STEM in India. Specifically, I propose an experiment with 610 students at a large engineering college, where students are randomly assigned to 203 project teams that vary on how leaders are selected within teams (peer-nominated vs. researcher-assigned leaders based on social skills). Using an incentivized app development competition addressing challenges in rural India, the study analyzes how examines leaders characteristics differ across leadership selection methods, and measures how these influence team dynamics and performance.
Last Published June 24, 2025 07:25 AM June 27, 2025 03:34 AM
Intervention (Public) The intervention involves a randomized controlled trial with student teams at an engineering college in rural India. Students are first randomly assigned into teams of three, stratified by academic year (Year 1, Year 2, Year 3), with each team containing one female and two males (Female-minority arm), OR two females and one male (Female-majority arm). These 203 teams are then randomly assigned across two factors: 1) Peer-Nominated Leader (T1): Team members discuss and nominate a leader from among themselves. 2) Researcher-Assigned Leader (T2): The researcher assigns the leader based on the social perceptiveness --specifically the highest PAGE (Perceived AI Generated Emotions) score among the team members. All teams participate in a two-week app concept competition focused on solving rural challenges in agriculture or education. The competition simulates real-world teamwork under pressure, allowing the study to measure how different methods of leader selection impact team dynamics and project outcomes The intervention involves a randomized controlled trial with student teams at an engineering college in rural India. Students are first randomly assigned into teams of 3 with variation in gender composition (female-majority or female minority team) and these teams are then randomly assigned across two arms: 1) Peer-Nominated Leader (T1): Team members discuss and nominate a leader from among themselves. 2) Researcher-Assigned Leader (T2): The researcher assigns the leader based on the social perceptiveness --specifically the highest PAGE (Perceived AI Generated Emotions) score among the team members. All teams participate in a two-week app concept competition focused on solving rural challenges in agriculture or education. The competition simulates real-world teamwork under pressure, allowing the study to measure how different methods of leader selection impact team dynamics and project outcomes.
Primary Outcomes (End Points) Outcome #1: Differences in leader characteristics between peer-nominated and researcher-assigned teams, and female-minority and female-majority teams Outcome # 2: Differences in average project rating Outcome # 3: Reported Team Effectiveness Outcome # 4: Reported Team Leader Effectiveness Outcome # 5: Perceived Influence Outcome #1: Differences in leader characteristics between peer-nominated and researcher-assigned teams Outcome # 2: Differences in average project rating Outcome # 3: Reported Team Effectiveness Outcome # 4: Reported Team Leader Effectiveness Outcome # 5: Perceived Influence
Primary Outcomes (Explanation) Outcome #1: Differences in leader characteristics between peer-nominated and researcher-assigned teams, and female-minority and female-majority teams Outcome # 2: Differences in average project rating (each project will be rated by 2 trained independent raters, based on a pre-defined rubric) across treatment arms Outcome # 3: Reported Team Effectiveness (index constructed from 4 questions in the endline) Outcome # 4: Reported Team Leader Effectiveness (reported on a scale of 1-10) Outcome # 5: Perceived Influence (probability of women team members being voted as most influential - Karpowitz et al. 2024) Outcome #1: Differences in leader characteristics between peer-nominated and researcher-assigned teams Outcome # 2: Differences in average project rating (each project will be rated by 2 trained independent raters, based on a pre-defined rubric) across treatment arms Outcome # 3: Reported Team Effectiveness (index constructed from 4 questions in the endline) Outcome # 4: Reported Team Leader Effectiveness (reported on a scale of 1-10) Outcome # 5: Perceived Influence (probability of women team members being voted as most influential - Karpowitz et al. 2024) - Only to test the effect of variation in gender composition of teams
Experimental Design (Public) This study uses a 2×2 factorial randomized controlled trial with 610 undergraduate students at an engineering college in rural India. Students are first randomly assigned into teams of three, stratified by academic year (Year 1, Year 2, Year 3), with each team containing one female and two males, OR two females and one male. These 203 teams are then randomly assigned across two factors: Factor 1 - Team Gender Composition: Female-majority teams (2 females, 1 male) vs Female-minority teams (1 female, 2 males) Factor 2 - Leadership Selection Method: 1) Peer-Nominated Leader (T1): Teams nominate a leader from among their members 2) Researcher-Assigned Leader (T2): The leader is assigned by the researcher based on the team member's score on a social perceptiveness assessment Teams participate in an incentivized app development competition addressing rural Indian challenges. The competition simulates real-world teamwork under pressure, allowing the study to measure how different methods of leader selection impact team dynamics and project outcomes. 610 undergraduate students at an engineering college in rural India are first randomly assigned into teams of three, stratified by academic year (Year 1, Year 2, Year 3), with each team containing one female and two males, OR two females and one male. These 203 teams are then randomly assigned across two arms: Peer-Nominated Leader (T1): Teams nominate a leader from among their members Researcher-Assigned Leader (T2): The leader is assigned by the researcher based on the team member's score on a social perceptiveness assessment -- specifically the highest PAGE (Perceived AI Generated Emotions) score among the team members. Teams participate in an incentivized app development competition addressing rural Indian challenges. The competition simulates real-world teamwork under pressure, allowing the study to measure how different methods of leader selection impact team dynamics and project outcomes.
Randomization Unit Two-stage randomization: 1) Individual students randomized into teams of 3 (stratified by academic year) 2) Teams randomized into 2×2 factorial treatment conditions (stratified by academic year and gender composition) Two-stage randomization: 1) Individual students randomized into teams of 3 (within academic year and such that each team as AT LEAST 1 female and 1 male member) 2) Teams randomized into leadership selection treatment conditions (stratified by academic year and gender composition)
Sample size (or number of clusters) by treatment arms 105 teams in Female-majority teams (across peer-nominated and researcher-assigned conditions) 98 teams in Female-minority teams (across peer-nominated and researcher-assigned conditions) 101 teams in Peer-Nominated Leader condition (across female-majority and female-minority teams) 102 teams in Researcher-Assigned Leader condition (across female-majority and female-minority teams) Total: 203 teams, 610 students (3 students per team) Design: 2×2 factorial examining main effects of (1) team gender composition and (2) leadership selection method 101 teams in Peer-Nominated Leader condition (across female-majority and female-minority teams) 102 teams in Researcher-Assigned Leader condition (across female-majority and female-minority teams) Total: 203 teams, 610 students (3 students per team)
Secondary Outcomes (End Points) Google Slides/Docs activity: Version history to view detailed contribution records WhatsApp Group activity Measures of productivity (number of meetings conducted, number of hours worked) Google Slides/Docs activity: Version history to view detailed contribution records WhatsApp Group activity Measure of Sexism (only to test the effect of variation in gender composition of teams) Measures of productivity (number of meetings conducted, number of hours worked)
Secondary Outcomes (Explanation) Google Slides/Docs activity: Version history to view detailed contribution records WhatsApp Group activity: constructed using activity data on team-level WhatsApp groups Measures of productivity (number of meetings conducted, number of hours worked) Google Slides/Docs activity: Version history to view detailed contribution records WhatsApp Group activity: constructed using activity data on team-level WhatsApp groups Measures of productivity (number of meetings conducted, number of hours worked) Measure of Sexism (Glick et. al, 1996 -- “The Ambivalent Sexism Inventory: Differentiating Hostile and Benevolent Sexism.”)
Back to top